---
title: "Covid19 Northern Ireland"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    vertical_layout: fill
    source_code: embed
    css: style.css
    social: menu
    theme:
      version: 4
      bg: "white"
      fg: "#202020"
      primary: "#2A9D8F"
      navbar-bg: "#2A9D8F"
      base_font: 
          google: Roboto
      heading_font:
          google: Roboto
---

```{css}

.value-output {
  color: white;
}

.caption {
  color: white;
}

```

```{r setup, include=FALSE}
library(flexdashboard)
library(dplyr)
library(zoo)
library(plotly)
### Pulling most recent data from Github
national_df <- readr::read_csv("https://raw.githubusercontent.com/obrienjoey/covid19northernireland/main/data/ni_covid_national.csv")
last_national_data <- national_df %>% na.omit() %>% filter(date == max(date))

local_df <- readr::read_csv("https://raw.githubusercontent.com/obrienjoey/covid19northernireland/main/data/ni_covid_local.csv")

inpatient_df <- readr::read_csv("https://raw.githubusercontent.com/obrienjoey/covid19northernireland/main/data/ni_inpatients.csv")

hosp_df <- readr::read_csv("https://raw.githubusercontent.com/obrienjoey/covid19northernireland/main/data/ni_covid_hospital.csv")

pal_navy <- '#264653'
pal_teal <- '#2A9D8F'
pal_yellow <- '#E9C46A'
pal_orange <- '#F4A261'
pal_red <- '#E76F51'
ablack <- '#202020'

shapefile <- sf::st_read(
  "data/shapefiles/OSNI_Open_Data_-_Largescale_Boundaries_-_Local_Government_Districts_(2012).shp")

```

Dashboard
=======================================================================

Row
-----------------------------------------------------------------------

### Cases {.value-box}
```{r}
valueBox(value = paste(format(last_national_data$cases, big.mark = ","), "", sep = " "), 
         caption = "Cases", 
         icon = "fas fa-plus-square", 
         color = pal_navy)
```

### Tests (last 5 min) {.value-box}
```{r}
valueBox(value = paste(format(last_national_data$tests, big.mark = ","), "", sep = " "), 
         caption = "Tests", 
         icon = "fas fa-vial", 
         color = pal_teal)
```

### Deaths {.value-box}
```{r}
valueBox(value = paste(format(last_national_data$deaths, big.mark = ","), "", sep = " "), 
         caption = "Deaths", 
         icon = "fas fa-window-close", 
         color = pal_yellow)
```

### Hospital Patients {.value-box}
```{r}
valueBox(value = paste(format(last_national_data$covid_patients, big.mark = ","), "", sep = " "), 
         caption = "Hospital Patients", 
         icon = "fas fa-hospital", 
         color = pal_orange)
```

### ICU Patients {.value-box}
```{r}
valueBox(value = paste(format(last_national_data$covid_ICU, big.mark = ","), "", sep = " "), 
         caption = "ICU Patients", 
         icon = "fas fa-procedures", 
         color = pal_red)
```

Row
-----------------------------------------------------------------------

### Covid19 Historical Data {data-width=600}
    
```{r}


national_df %>%
  select(date, cases, covid_patients, covid_ICU) %>%
  unique() %>%
  mutate(cases_ma7 = round(rollmean(x = cases, # column to take
                              k = 7, # rolling time period
                              align = "right", #leave values above the top
                              fill = NA),2)) %>%
  plotly::plot_ly() %>%
  plotly::add_bars(x = ~date,
                   y = ~cases,
                   color = I(pal_teal),
                   hoverinfo = 'text',
                   text = ~paste(format(date, '%d-%b-%Y'), '

Cases:', cases),
                   name = 'Cases') %>%
  plotly::add_lines(x = ~date,
                    y = ~cases_ma7,
                    name = '7-day Average',
                    mode = 'lines',
                    hoverinfo = 'text',
                    text = ~paste(format(date, '%d-%b-%Y'), '

7-day Average:', cases_ma7),
                    color = I(ablack)) %>%
  plotly::add_bars(x = ~date,
                   y = ~-covid_patients,
                   name = 'Hospital',
                   hoverinfo = 'text',
                   text = ~paste(format(date, '%d-%b-%Y'), '

Hospital:', covid_patients),
                   color = I(pal_orange)) %>%
  plotly::add_bars(x = ~date,
                   y = ~-covid_ICU,
                   name = 'ICU',
                   hoverinfo = 'text',
                   text = ~paste(format(date, '%d-%b-%Y'), '

ICU:', covid_ICU),
                   color = I(pal_red)) %>%
  plotly::layout(barmode='overlay',
                 yaxis = list(title = '',
                              tickvals = seq(-1000,2000,500), 
                              ticktext = c('1000','500','0','500','1000','1500','2000')),
                 xaxis = list(title = 'Date',
                              tickformat = "%b %Y"),
                 legend = list(x = 0.05, y = 0.9),
                 hovermode = 'text')

```
   
### Deaths and Positivity Rate {data-width=400}

```{r}

min_death_date <- national_df %>%
  select(date, deaths) %>%
  na.omit() %>%
  pull(date)

death_plot <- national_df %>%
  select(date, deaths) %>%
  na.omit() %>%
  plotly::plot_ly() %>%
  plotly::add_markers(x = ~date,
                    y = ~deaths,
                    mode = 'lines',
                    hoverinfo = 'text',
                    text = ~paste(paste(format(date, '%d-%b-%Y'), '

Deaths:', deaths)),
                                  color = I(pal_yellow)) %>%
  plotly::layout(barmode='overlay',
                 yaxis = list(title = 'Deaths'),
                 xaxis = list(title = '',
                              range = c(min_death_date - 1, max(national_df$date) + 1)),
                 hovermode = 'text')

test_plot <- national_df %>%
  select(date, cases, tests) %>%
  unique() %>%
  filter(tests > 0) %>%
  mutate(negative = tests - cases) %>%
  plotly::plot_ly(type = 'scatter', 
                  mode = 'none', 
                  stackgroup = 'one', 
                  groupnorm = 'percent') %>%
  plotly::add_trace(x = ~date, 
                    y = ~cases, 
                    fillcolor = pal_navy,
                    hoverinfo = 'text',
                    text = ~paste0('Positive : ', round(100*cases/tests,2), '% (', cases, ')')) %>%
  plotly::add_trace(x = ~date,
                    y = ~negative,
                    fillcolor = pal_teal,
                    hoverinfo = 'text',
                    text = ~paste0('Negative : ', round(100*negative/tests,2), '% (', negative, ')')) %>%
  plotly::layout(barmode='overlay',
                 yaxis = list(title = 'Fraction of Tests',
                              showgrid = FALSE,
                              ticksuffix = '%'),
                 xaxis = list(title = 'Date',
                              tickformat = "%b %Y",
                              showgrid = FALSE),
                 hovermode = "x unified",
                 showlegend = FALSE)

subplot(death_plot, test_plot, nrows = 2,
        titleX = TRUE, titleY = TRUE) 

```   

Regional Data
=======================================================================

### Number of Cases by Electoral District (`r format(max(local_df$date), '%d-%b-%Y')`) { data-width=500 }
```{r}

local_df_summary <- local_df %>%
              unique() %>%
              mutate_if(is.numeric, ~replace(., is.na(.), 0)) %>%
              group_by(area) %>%
              summarise(tests_1 = sum(tail(tests,1)),
                        tests_7 = sum(tail(tests,7)),
                        tests_30 = sum(tail(tests,30)),
                        tests_all = sum(tests),
                        cases_1 = sum(tail(cases,1)),
                        cases_7 = sum(tail(cases,7)),
                        cases_30 = sum(tail(cases,30)),
                        cases_all = sum(cases),
                        deaths_1 = sum(tail(deaths,1)),
                        deaths_7 = sum(tail(deaths,7)),
                        deaths_30 = sum(tail(deaths,30)),
                        deaths_all = sum(deaths)) %>%
              janitor::adorn_totals("row") %>%
              as_tibble() %>%
              filter(area != 'Missing Postcode') %>%
              na.omit() %>%
              tidyr::pivot_longer(cols = tests_1:deaths_all,
                           names_to = 'category') %>%
              tidyr::separate(category, 
                       into = c("Category", "Time"),
                       sep="_(?=[^_]+$)")

map_df <- local_df_summary %>%
  filter(Time == 1,
         Category == 'cases') %>%
  inner_join(shapefile, .,
             by = c('LGDNAME' = 'area'))

mapview::mapview(sf::st_zm(map_df), 
                 zcol = c('value'),
                 layer.name = 'Cases')

```

### Summary of Cases (Deaths) by Electoral District (as of `r format(max(local_df$date), '%d-%b-%Y')`)  { data-width=400 }

```{r}

local_df %>%
  unique() %>%
  mutate_if(is.numeric, ~replace(., is.na(.), 0)) %>%
  group_by(area) %>%
  summarise(cases_1 = sum(tail(cases,1)),
            cases_7 = sum(tail(cases,7)),
            cases_30 = sum(tail(cases,30)),
            cases_all = sum(cases),
            deaths_1 = sum(tail(deaths,1)),
            deaths_7 = sum(tail(deaths,7)),
            deaths_30 = sum(tail(deaths,30)),
            deaths_all = sum(deaths)) %>%
  janitor::adorn_totals("row") %>%
  as_tibble() %>%
  na.omit() %>%
  filter(area != 'Missing Postcode') %>%
  mutate('1 Day' = paste0(format(cases_1, big.mark = ',', trim = TRUE),
                          ' (', format(deaths_1, big.mark = ',', trim = TRUE), ')'),
         '7 Days' = paste0(format(cases_7, big.mark = ',', trim = TRUE),
                           ' (', format(deaths_7, big.mark = ',', trim = TRUE), ')'),
         '30 Days' = paste0(format(cases_30, big.mark = ',', trim = TRUE),
                            ' (', format(deaths_30, big.mark = ',', trim = TRUE), ')'),
         'All Time' = paste0(format(cases_all, big.mark = ',', trim = TRUE),
                             ' (', format(deaths_all, big.mark = ',', trim = TRUE), ')')) %>%
  select('Area' = area,
         '1 Day',
         '7 Days',
         '30 Days',
         'All Time') %>%
  DT::datatable(rownames = FALSE,
                fillContainer = T,
                options = list(pageLength = nrow(.),
                               searching = FALSE,
                               dom = 't',
                               ordering=F,
                               scrollY = 450,
                columnDefs = 
                           list(list(className = 'dt-left', targets = 0),
                                list(className = 'dt-center', targets = 1:4)))) %>%
  DT::formatStyle(
      'Area',
      target = "row",
      fontWeight = DT::styleEqual('Total', "bold")
    )
  
```

Hospital Data
=======================================================================

### ICU Facilities (`r format(max(local_df$date), '%d-%b-%Y')`) { data-width=500 }

```{r}

national_df %>%
  select(date, covid_ICU, non_covid_ICU, unoccupied_ICU_beds) %>%
  na.omit() %>%
  tidyr::pivot_longer(cols = covid_ICU:unoccupied_ICU_beds,
               names_to = 'group',
               values_to = 'number') %>%
  group_by(date) %>%
  mutate(prop = number/sum(number),
         group = case_when(group == 'covid_ICU' ~ 'Covid Patients',
                           group == 'non_covid_ICU' ~ 'Non-Covid Patients',
                           group == 'unoccupied_ICU_beds' ~ 'Unoccupied')) %>%
  plotly::plot_ly(x = ~date, y = ~round(prop * 100,2), color = ~group,
          hoverinfo = 'text', colors = c(pal_orange,pal_teal,pal_yellow),
          text = ~paste0(group, ': ', number, ' (', round(prop * 100,2), '%)')
          ) %>%
  plotly::add_bars() %>%
  plotly::layout(barmode = 'stack',
         yaxis = list(title = 'Fraction of ICU Beds',
                      showgrid = FALSE,
                      ticksuffix = '%'),
         xaxis = list(title = 'Date',
                      tickformat = "%b %Y",
                      showgrid = FALSE,
                      hoverformat = "%b %d, %Y"),
         hovermode = "x unified",
         showlegend = FALSE)


```

### Inpatient Data (`r format(max(local_df$date), '%d-%b-%Y')`) { data-width=500 }

``` {r}

inpatient_df %>%
  filter(gender == 'All') %>%
  group_by(date, age_band) %>%
  summarise(covid_patients = sum(covid_patients)) %>%
  mutate(age_band = gsub("Aged ", "", age_band)) %>%
  plot_ly(type = 'scatter', x = ~date, y = ~covid_patients, 
          color = ~age_band, mode = 'none', 
          stackgroup = 'one') %>%
  layout(barmode = 'stack',
         yaxis = list(title = 'Inpatients at Midnight'),
         xaxis = list(title = 'Date',
                      tickformat = "%b %Y",
                      showgrid = FALSE,
                      hoverformat = "%b %d, %Y"),
         hovermode = "x unified",
         legend = list(orientation = "h", 
                       xanchor = "center",  
                       x = 0.5,
                       y = 1.1))

```

### Total HSC Trust Inpatients (`r format(max(local_df$date), '%d-%b-%Y')`) { data-width=200 }

```{r}

hosp_df %>%
  group_by(hsc_trust) %>%
  summarise(admissions = sum(number_of_admissions)) %>%
  arrange(desc(admissions)) %>%
  plot_ly(x = ~ admissions, y = ~reorder(hsc_trust,admissions), color = I(pal_teal)) %>%
  add_bars() %>%
  layout(yaxis = list(title = 'HSC Trust',
                      showgrid = FALSE),
         xaxis = list(title = ''),
         showlegend = FALSE) %>%
  add_annotations(text = ~admissions,
                  y = ~reorder(hsc_trust,admissions),
                  x = ~admissions/2,
                  xref = "x",
                  yref = "y",
                  font = list(
                              size = 14,
                              color = 'white'),
                  showarrow = FALSE)
```

About
=======================================================================

**Covid 19 Northern Ireland Dashboard**

This dashboard provides an overview of the latest data regarding the 2019 Novel Coronavirus COVID-19 (2019-nCoV) outbreak in Northern Ireland. The data is updated every day to incorporate the most recent statistics via an additional piece of software hosted [here](https://github.com/obrienjoey/covid19northernireland). The dashboard itself is implemented in the `R` programming language, specifically using the `flexdashboard` package within an `Rmarkdown`framework. 

**Packages**

* Dashboard interface - the [`flexdashboard`](https://rmarkdown.rstudio.com/flexdashboard/) package. 
* Visualization - the [`plotly`](https://plot.ly/r/) package for the plots and [`mapview`](https://r-spatial.github.io/mapview/) package for the map.
* Data manipulation - [`dplyr`](https://dplyr.tidyverse.org/).
* Tables - the [`DT`](https://rstudio.github.io/DT/) package.

**Contact**

For further details feel free to check out the [Github repo](https://github.com/obrienjoey/covidni_dashboard) or contact me via [Twitter](https://twitter.com/obrienj_).